  Xizer
join:2004-02-05
New York, NY
 ·Cox HSI
 ·AT&T U-Verse
1 edit | So much wasted potential bandwidth usage AT&T must be building this backbone to benefit FiOS customers since AT&T's own subscribers now have bandwidth caps that effectively limit them to using their connection at an average speed of 768 Kbps... I don't see terabit backbones getting saturated by AT&T customers any time soon. | |
|
 |  Mike_
join:2003-06-24
Fort Lauderdale, FL
 ·Verizon Wireless..
·Comcast
| Re: So much wasted potential bandwidth usage This is true. Really they're only benefiting business ethernet lines and the like, as well as competitors. Residential will see nothing new, other than maybe new hops/routes. If Comcast really is also doing something, it will definitely improve and solidify the D3 and future D3 areas preventing any congestion.
--
If it ain't broke, don't fix it. Unless you know what you're doing. | |
|
| | timing? Comcast and AT&T want a bigger share of the wireless market, so backbone upgrades seem aimed more to that goal than actually doing more for the wired customer. | |
|
 mix
join:2002-03-19
Utica, MI | Upgrades Are all these upgrades new fiber or are they figuring out how to squeeze more throughput out of existing networks? | |
|
| | | Re: Upgrades said by mix:Are all these upgrades new fiber or are they figuring out how to squeeze more throughput out of existing networks?
There is less free than there used to be, but there still is a lot of dark fiber buried along railroad rights of way that hasn't been lit yet. Other than some equipment on the segments, the cost of lighting up more fiber is pretty cheap.
And yes, equipment upgrades can get more bandwidth out of already existing and used lit fiber. | |
|
| | Reviews:
·Optimum Online
·Verizon FiOS
| Re: Upgrades said by fAcEtIOUs:said by mix:Are all these upgrades new fiber or are they figuring out how to squeeze more throughput out of existing networks?
There is less free than there used to be, but there still is a lot of dark fiber buried along railroad rights of way that hasn't been lit yet. Other than some equipment on the segments, the cost of lighting up more fiber is pretty cheap. And yes, equipment upgrades can get more bandwidth out of already existing and used lit fiber. On the other hand, it's just cheaper for carriers to throttle your long-distance internet routes to save money as they've been doing since broadband beyond 10mbit days (both in the downstream and upstream tiers, if you ever got upload beyond 10mbit.. most dont' have it even today). UPTO and NOT GUARANTEED still keep them 'off the hook' for being cheap on backbone access, read your TOS! | |
|
|  espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2
·Clear Wireless
| said by mix:Are all these upgrades new fiber or are they figuring out how to squeeze more throughput out of existing networks? The IEEE 802.3ba standard for 40/100GigE interfaces was just finalized last June, so the product based on that standard is just starting to come to the market today.
All of this is over existing fiber, and allows for greater efficiency than trying to do src/dst IP load balancing across multiple 10gig links. | |
|
| iansltx
join:2007-02-19
Golden, CO
kudos:2 | Mostly upgrading current strands...so you have 100G of capacity where there was 40G before. | |
|
| | | | Re: Upgrades They are using 100G per wavelength not fiber. Verizon was using 44 channel ROADMs the last I knew and there were looking to upgrade to 88 channel ROADMs. This would be 88x100G circuits per fiber XMT/RCV pair. | |
|
| | | iansltx
join:2007-02-19
Golden, CO
kudos:2 | Re: Upgrades You are correct. Gotta love DWDM. | |
|
|  dvd536as Mr. Pink as they comePremium
join:2001-04-27
Phoenix, AZ
kudos:4 | said by mix:Are all these upgrades new fiber or are they figuring out how to squeeze more throughput out of existing networks?
If theres a way T can figure out how to get something for nothing you bet thats whats going on!
--
The early bird gets the worm but the second mouse gets the cheese | |
|
 IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Useless... Pointless...
Comcast is all about getting you to your cap faster and faster. And wait until overages come... Oh yes they will. Anything to monetize what's turning into a dumb pipe.
The Comcast cap has been at the same level since October 2008. What a joke.
The cap was never about addressing congestion. There's already a congestion management system for that.
Not fooling me.
--
"We're going to start at one end of (Fallujah), and we're not going to stop until we get to the other. If there's anybody left when that happens, we're going to turn around and we're going to go back and finish it."
Lt. Col. Pete Newell: 1st Inf. US Army | |
|
| |
|  AlcoholPremium
join:2003-05-26
Climax, MI
kudos:3 | Same can be said about AT&T. | |
|
| dvd536as Mr. Pink as they comePremium
join:2001-04-27
Phoenix, AZ
kudos:4 | said by IPPlanMan:Pointless...
Comcast is all about getting you to your cap faster and faster. And wait until overages come... Oh yes they will. Anything to monetize what's turning into a dumb pipe.
The Comcast cap has been at the same level since October 2008. What a joke.
The cap was never about addressing congestion. There's already a congestion management system for that.
Caps are all about protecting LUCRATIVE VIDEO! always have.
--
The early bird gets the worm but the second mouse gets the cheese | |
|
| | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Re: Useless... Exactly... Comcast wants TV for TV and Internet for Internet....
Internet for TV? That's a no-no...
Netflix, iTunes/Apple TV, Amazon, Vudu?
All of these alternatives scare the hell out of Comcast. | |
|
| | | said by IPPlanMan:Pointless...
Comcast is all about getting you to your cap faster and faster. And wait until overages come... Oh yes they will. Anything to monetize what's turning into a dumb pipe.
The Comcast cap has been at the same level since October 2008. What a joke.
The cap was never about addressing congestion. There's already a congestion management system for that.
Not fooling me.
I don't follow your logic. Are you saying it is pointless to upgrade the network to handle the traffic growth for all? Would you would rather cap the network growth and let congestion management handle it from there for everyone vs the cap for the extreme 1% minority? | |
|
| | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Re: Useless... I'm saying to dump the cap, admit that it wasn't about preventing congestion but protecting Comcast's premium channel services and HD on Demand, and let the congestion management system handle it.
This so called network growth isn't going to make the cap go away. At some point in the future, we'll just get to the cap faster... And then: Hello Overages! So what's the point of that?
This is called market conditioning. Comcast is trying to make us adjust to the new normal...
Not fooling me Comcast.
--
"We're going to start at one end of (Fallujah), and we're not going to stop until we get to the other. If there's anybody left when that happens, we're going to turn around and we're going to go back and finish it."
Lt. Col. Pete Newell: 1st Inf. US Army | |
|
| | | | | Re: Useless... I still don't understand what your point has to do with this article. Upgrading the core network to support growth across ALL users is a good thing for everyone.
This really has nothing to do with the "CAP Brigade" or 1% of users who believe they are entitled to unlimited bandwidth at the same flat rate as the 99% of others. While I understand you want your costs for year over year growth to be spread to all users, everyone benefits from the overall network upgrades. | |
|
| | | | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1
1 edit | Re: Useless... said by yt:While I understand you want your costs for year over year growth to be spread to all users, everyone benefits from the overall network upgrades.
Since the Oct 2008 cap (which has been at the same level since), my speeds have been the same, and the cost has gone up substantially.
Please let me know when that benefit occurs...
--
"We're going to start at one end of (Fallujah), and we're not going to stop until we get to the other. If there's anybody left when that happens, we're going to turn around and we're going to go back and finish it."
Lt. Col. Pete Newell: 1st Inf. US Army | |
|
| | Who cares with CAPS! Great so with CAPS I'll be able to download for a whole minute before being disconnected or being charged $100s of dollars! | |
|
| | | Re: Who cares with CAPS! you didnt read... 
This is backbone not last mile.  | |
|
| | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Re: Who cares with CAPS! So what problem is this so called "investment" solving then?
If it doesn't raise Comcast's cap, raise its speed, or lower the price, what exactly are we supposed to feel good about and how does it benefit anyone?
This sounds like a whole lot of useless hooey...
--
"We're going to start at one end of (Fallujah), and we're not going to stop until we get to the other. If there's anybody left when that happens, we're going to turn around and we're going to go back and finish it."
Lt. Col. Pete Newell: 1st Inf. US Army | |
|
| | | espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2 Reviews:
·Clear Wireless
| Re: Who cares with CAPS! said by IPPlanMan:So what problem is this so called "investment" solving then? It solves having to scale the current core of CRS-1 gear by more expensive channelized 40Gig SoNET interfaces. Since Ethernet has far simpler queuing and timing demands, you can use the cheaper MSCs to support an 802.3ba-based PLIM.
The net effect is:
1) Vast reduction in costs to scale the environment because the interface cost is drastically reduced over SoNET-based channelized interfaces. (ie, OC768)
2) Reduction of recurring maintenance costs due to using lower cost hardware.
3) Simplification of infrastructure -- fewer links, simpler interfaces, less complex configuration. Reduces MTTR because problems are easier to isolate and resolve.
4) Better efficiency on inter-switch links under heavy load. As utilization increases, static source/destination IP hashing algorithms lead to an inefficient distribution of traffic. For example, across 4 links the load distribution rarely works out to 25/25/25/25. Instead, you end up with something like 40/30/25/5 so you end up massively overbuilding inter-switch links to account for inadequacies in the hashing method.
said by IPPlanMan:This sounds like a whole lot of useless hooey... It's statements like these where it becomes abundantly clear who actually works on large networks, and who just stayed in a Holiday Inn Express last night. | |
|
| | | | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1
1 edit | Re: Who cares with CAPS! That's a great explanation, but so what? It doesn't help me or a lot of other people for that matter. It's not making my cap any higher than it was almost 3 years ago. My internet/cable bill is going up and up and up, just to keep what I have now. So much for those alleged cost-savings being passed onto the consumer.
You can denigrate and minimize my statements all you want, but this is what Comcast's so-called investments on behalf of the residential customer amount to:
»www.youtube.com/watch?v=Ug75diEyiA0
I'm not the only one who feels this.
--
"We're going to start at one end of (Fallujah), and we're not going to stop until we get to the other. If there's anybody left when that happens, we're going to turn around and we're going to go back and finish it."
Lt. Col. Pete Newell: 1st Inf. US Army | |
|
| | | | | espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2 Reviews:
·Clear Wireless
| Re: Who cares with CAPS! said by IPPlanMan:That's a great explanation, but so what? It doesn't help me or a lot of other people for that matter. The irony is that the mantra around here is "upgrade the damn network!" -- but now that there is an article posted about providers upgrading their network, the immediate reaction is "this is worthless because it doesn't immediately benefit me!"
Capacity doesn't magically appear -- it takes careful, deliberate planning. This gear doesn't just show up on the dock and install itself; someone has to spec the gear, develop a plan, and then complete tasks that are akin to changing the tires on a moving car. Everybody wants more capacity, nobody wants an outage.
This is the stuff that eventually will lead to more capacity across the system. | |
|
| | | | | |  EnasYorlThieves World
join:2001-12-02
West | Re: Who cares with CAPS! said by espaeth  ...
Capacity doesn't magically appear -- it takes careful, deliberate planning. This gear doesn't just show up on the dock and install itself; someone has to spec the gear, develop a plan, and then complete tasks that are akin to changing the tires on a moving car. Everybody wants more capacity, nobody wants an outage.
This is the stuff that eventually will lead to more capacity across the system.
| |
|
| | | | brad
join:2007-09-06
Etobicoke, ON | said by espaeth:said by IPPlanMan:This sounds like a whole lot of useless hooey... It's statements like these where it becomes abundantly clear who actually works on large networks, and who just stayed in a Holiday Inn Express last night. Forget trying to explain this to the overwhelming majority of the idiots on DSLR.
and Holiday Inn is too high class for this guy. | |
|
| | | |  lovswr
join:2001-09-15
Smyrna, GA | espaeth I disagree with your core contention that the SONET gear is eliminated. The long haul is still SONET (as you state 0c-768). There is no multi-hundred fiber link carrying straight Ethernet (Layer-2?) & nothing else. SONET is (& for the foreseable future?) the Layer 1/physical, long haul, transport system. All this does is push the muxing one order higher (i.e. 40g presentation to the SONET equipment vs. 100g presentation). In that case, I do agree that there will be savings, but mayhaps, not as much as you may suggest.
Ethernet is really (getting to be) ubiquitous, but there are still no good mechanism for Layer 1 fault recovery other than multiple links/sheer B/W.
--
lovswr = good hivswr = bad | |
|
| | | | | espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2 Reviews:
·Clear Wireless
| Re: Who cares with CAPS! said by lovswr:The long haul is still SONET (as you state 0c-768). There is no multi-hundred fiber link carrying straight Ethernet (Layer-2?) & nothing else. SONET is (& for the foreseable future?) the Layer 1/physical, long haul, transport system. No need for SONET when you can do native transport straight into DWDM. Several vendors offer ROADM gear with native 100GigE interfaces either shipping currently or shipping soon. Ciena, Cisco, Tellabs and Fujitsu all have gear with this capability.
That works great if you have your own dark fiber between locations like these carriers do, but even if you're in private enterprise there are options to purchase wavelength transport (ie, Qwest's QWave service) and you can still do native Ethernet transport across the country.
said by lovswr:Ethernet is really (getting to be) ubiquitous, but there are still no good mechanism for Layer 1 fault recovery other than multiple links/sheer B/W. Most of these networks aren't relying on any kind of Layer 1 protect services -- Layer 3 already handles that nicely through dynamic routing protocols with multiple paths. | |
|
| | |
| | | | | Re: Who cares with CAPS! i can agree
But i think its to say that they have a faster connection than the other guy. | |
|
| | | IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Couldn't have said it any better... | |
|
| | | EnasYorlThieves World
join:2001-12-02
West | said by Xizer
Pointless to upgrade the backbone if no one at the last mile is allowed to use barely more than a dial-up connection's worth of bandwidth in a month.
The problem is that there is no banks to spill the water aka data over the riverbanks to keep the water moving so it slows down and backs up.
You have to lay the foundation to build a house. | |
|
| IPPlanManHoly Cable Modem Batman
join:2000-09-20
Washington, DC
kudos:1 | Yup. Exactly right. | |
|
decifal
join:2007-03-10
Bon Aqua, TN
kudos:1 | why Why bother | |
|
| Reviews:
·Verizon FiOS
| Re: why why not? if they have the capacity, they can start offering higher end-user speeds once everything is in place. do you want them to deliver a 10Gb link to your home when the backbone only supports 12Gb total? i'm sure higher caps will come of this too. everyone will be asking the question, why bother upgrading bandwidth if you're going to limit our use further? no need to even cap if there's no longer congestion. | |
|
| | | | Re: why Caps are not really about congestion, and they will never be raised. Besides increasing network efficiency for ISPs, capacity will be used to reduce over-subscription a little as new customers are added and old customers start to actually use the connections they have overpaid for for years, and possibly the new capacity will allow new expensive speed tiers for suckers who like to show off their speedtest results. | |
|
| | | | | Re: why Think long term and not short term, amigo. Then, you actually start to comprehend what's involved. | |
|
|
|
